[gdw]
This lexeme is unique to Anglo-Norman, and its origin is problematic: it appears to be the result of a semantic conflation of rescuns (‘sunset’) with the form of rescus1 (‘rescue’). A straightforward denasalisation of rescuns to rescus may have been possible (cf. FEW abscondere 24,50b and 51a, by analogy with ponere vs. positus, 51b.n1). Alternatively, it may be that forms like recous and rescous were originally misinterpretations of recons and rescons, with u and n often indistinguishable in manuscripts (cf. the variant readings in the triv and Bor Cust attestations). Once read as -ous, the word for ‘sunset’ may have been understood as a sense of rescus1, allowing all other variant spellings of that lexeme, including those with an r, to feature – unetymologically, unsemantically and ungrammatically – in the sense of ‘sunset’.
A third association is possible, this time with recurs1 (‘return (from one place to another’), and its stem cours, which has the astronomical sense ‘course (of a heavenly body)’. However, not all variant spellings of rescours1 can be interpreted as recurs1.
Consequently, this word, with the sense of ‘sunset’ has been given its own AND entry, as a conflation of possibly three etymologies/semantics.